FIDE's President Is Arkady Dvorkovich, Not Timur Turlov: Inside a FRITZ 20 Promotional Piece

On 30 August 2026, a reader in Kazan sent me a link with three lines of text: "Check this for me — has FIDE changed presidents?" The headline stated that Timur Turlov — described as FIDE president — had just met the President of the Republic of Cyprus, Nikos Christodoulides, in Nicosia. I opened the official website of the International Chess Federation. It took forty seconds. The portrait on the right column was still Arkady Dvorkovich, elected at the Batumi Congress in October 2026 and re-elected in Chennai in August 2026. No handover communiqué, no extraordinary congress, no agenda item bearing Turlov's name. The body of the article was given over almost entirely to FRITZ 20, a chess program, under three bold taglines: your personal chess trainer, your toughest opponent, your strongest ally.

Three pieces — a political headline, a name attached to the wrong office, a block of advertising copy — do not sit next to each other by accident. They sit next to each other when a writer needs an authority hook to sell something. I have read thousands of chess articles across more than five decades, and I have learned one simple thing: when an article opens with a title of office and closes with a product, the middle is usually empty. That emptiness is what I want to measure here.

FIDE was founded in Paris in 2026 and administers the Elo system, the laws of the game and the world championship cycle. Its president is elected by a congress of member federations for a four-year term. In 2026, Dvorkovich — a former Russian deputy prime minister, then chairman of the board of Russian Railways — defeated Nigel Short in Batumi. In 2026, in Chennai, he was re-elected over Andrii Baryshpolets by a wide margin. That record is public, minuted, dated and delegate-listed.

Timur Turlov is someone else entirely. Born in Kazakhstan in 2026, he founded and runs Freedom Holding Corp, a financial group listed on Nasdaq under the ticker FRHC and headquartered in Almaty. His name appears in Kazakh chess as a sponsor and promoter of major events. A generous sponsor of chess is normal and welcome. A sponsor described as FIDE president is something else, and that something else is where the article begins to slide.

FRITZ, by contrast, has a real history, longer than most readers of a promotional piece would imagine. Fritz appeared in 2026, developed by Frans Morsch and Mathias Feist, inside the ChessBase product ecosystem. In October 2026, Deep Fritz drew 4-4 with Vladimir Kramnik over eight games in Manama, Bahrain. In November and December 2026, Deep Fritz beat Kramnik 4-2 in Bonn. Those are milestones with minutes, games and move-by-move commentary. A brand built on matches like that has earned the right to pride. It has not earned an exemption from verification.

So what belongs in a credible engine release note? An Elo estimate on independent engine rating lists, nodes per second on a specified hardware configuration, the evaluation architecture — classical or NNUE neural network, search depth at fixed time, average centipawn loss against a reference baseline, access to seven-piece Syzygy tablebases, scaling as thread count rises, and the opening book format shipped alongside. That is the vocabulary of the trade. The article I read did not contain a single line of it.

The absence is meaningful. A technical product that publishes no specifications is usually selling something other than strength. In computer chess, strength has become close to a free good: Stockfish is open source, Leela Chess Zero is open source, and both outplay every human and most commercial engines. If you want strength, you download it and pay nothing. So what are you paying for? The interface layer, the game database, the opening tree, the training mode, the way the program explains a bad move to a beginner.

That is a real market and not a small one. ChessBase has sold databases and interfaces to professional analysts for three decades. But precisely because the real product sits in the teaching layer, it is the layer hardest to verify. You can measure nodes per second in thirty seconds. You cannot measure "becomes a better teacher" in thirty seconds. Where measurement is hard, advertising inflates.

The recent history of this very ecosystem shows what happens when the technical layer is hidden. In February 2026, ChessBase released Fat Fritz 2. Within days, the community found the engine was built on Stockfish source code, released under the GPL. The Hamburg Regional Court issued an injunction against distribution, and Stockfish prevailed in the dispute. The lesson is not about who was right in a lawsuit. The lesson is that a product marketed on brand reputation, with its technical core undisclosed, was stripped bare by the community within days.

The real technological leap of the past decade did not come out of any boxed product either. NNUE — an efficiently updatable neural network — was published by Yu Nasu in 2026, and Stockfish integrated it in 2026. An academic idea, an open community, a jump in strength. No advertising campaign stood in the middle. In computer chess, the genuine turning point usually arrives first and the brand label is printed afterwards.

I came to chess from football and brought a working habit with me. People watch the players run. I watch the whole shape shift. On the board it is the same: people watch the fine move, I watch the space that move opens or closes. In this article, the largest space sits between the headline and the body. The headline speaks of institutional power. The body speaks of a software package. No bridge connects them, and the missing bridge is more worth analysing than either end.

I rebuilt the article's structure along a timeline, the way I once rebuilt matches. Opening: a meeting in Nicosia. Middle: adjectives about a product. Close: an invitation to buy. No passage explains why a head-of-state meeting relates to a chess engine. In match analysis I call this an empty block: the shape stands correctly on the diagram, but nobody owns the space in the middle. An empty block does not lose immediately. It loses when the ball arrives.

Three things in that article can be checked in ten minutes. One: the name of FIDE's head, on the official site, with dates. Two: Turlov's record, in corporate filings and financial press, with clear job titles. Three: the Fritz brand history, in game archives, with dates and scores. All three are public. An article that gets all three wrong, or gets one wrong and skips two, tells the reader more about its production process than about anything it says on chess.

I do not think the writer was malicious. I think the writer had a deadline. A headline with a proper name and a grand title draws clicks; a software advert pays per conversion; and in between sits a gap nobody checks because nobody is paid to check it. Systems do not lie, but they can only be heard when the data is thick enough. Here the data is so thin that its thinness is itself the finding.

But stopping at a corrected headline would miss the harder part. Why did that headline work? It worked because FIDE's governance story is genuinely opaque to the general public. An organisation that runs a global sport, elects a president every four years through national federations, holds a budget and deals with governments barely appears on the front page of any general sports outlet. When a subject is opaque, the gap gets filled with whatever sounds plausible. A young billionaire with money, chess connections and a name on big events is the perfect candidate for an office whose occupant most people cannot name.

The second harder part: buyers of chess software have no instrument for judging chess software. Independent engine rating lists exist, but they measure machine-versus-machine strength at long time controls — a figure close to meaningless for someone drilling endgames twenty minutes an evening. Nobody rates the quality of explanation. So the market pays for perceived authority, and perceived authority is built with headlines. As long as the measurement is missing, the advertising claim remains the currency.

I have been wrong in exactly this way once, and I remember it. In 2026, aged sixty, I wrote an analysis of RB Leipzig's gegenpressing based on xG data from 34 Bundesliga matchdays. The Russian online community called me a reactionary. I did not argue. I spent six weeks re-watching footage, logged 412 failed pressing situations, and published a correction with the numbers. The first piece got hammered. Data never takes offence. Since then I keep one rule: every claim needs a number or a timestamp behind it, and if there is none, I label it a hypothesis.

Applying that rule to FRITZ 20, here is what I can say without inventing anything. The Fritz brand has a real competitive record from 2026 through the Kramnik matches of 2026 and 2026. It belongs to the ChessBase ecosystem, a vendor with a long product history and also a history of licensing disputes. The three taglines in the article are market positioning, not specifications. A product positioned for both beginners and tournament players is a sensible commercial move, and it says nothing about training efficacy.

There is one contrarian point I believe more strongly: the best chess software of the next decade will not win on strength but on the ability to say no. A strong engine always points at the best move; a good teacher must know which move suits the student's level and which move teaches which lesson. That is a pedagogical problem, not a search problem, and the industry has barely begun to measure it. If FRITZ 20 genuinely carries an advance, it lives there — and that is precisely where the promotional copy offers no proof.

For young players, the gap is more dangerous than it looks. At the under-18 academies I have sat in, coaches under results pressure reach for fitness work and memorised openings instead of foundational technique. A program advertised as a training revolution without data accelerates that drift, because it gives the buyer the feeling that the technical problem has been solved with one click. I have seen the same pattern in youth football: buying equipment, not a curriculum.

On FIDE's side, the Turlov story should not be closed with a single denial. A major sponsor of Kazakh chess with ties to international events is worth tracking, and a global federation that needs money is worth tracking more. If his name appears in an official FIDE capacity, the communiqué will carry a date and a title. Until then, attaching the presidency to him is an error, and an error repeated often enough becomes an assumption. Assumptions are the dirtiest data there is, because nobody re-checks them.

Three things I will check over the next thirty days, and you can check with me: which title FIDE's official communications use for Turlov; whether independent engine rating lists record FRITZ 20 with a specified configuration and time control; and whether any review states which source code base the engine derives from. Three questions, three sources, no expertise required. I am sixty-nine. I still learn from the young. Football does not retire, and chess retires even less.

A false claim about FIDE's president can be corrected in a day. A market with no yardstick for training products takes years. What I want to leave behind is not a rebuke but a habit: every time you read a chess article that opens with a title of office and closes with a product, look for the middle. If the middle is empty, you already know what you are reading. And if you still buy, buy for the interface, the database, the curriculum — the things you can see — not for a headline.